highhash 0.2.0

High-speed hashing algorithms for Rust.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208

use std::hash::{BuildHasher, Hasher};

/// Hasher for MurmurHash3 32-bit implementation of the 128-bit hashing algorithm.
#[derive(Default)]
pub struct Murmur3Hasher128 {
    buffer: Vec<u8>,
}

impl Hasher for Murmur3Hasher128 {
    fn write(&mut self, bytes: &[u8]) {
        self.buffer.extend(bytes)
    }

    fn finish(&self) -> u64 {
        hash128(&self.buffer) as u64
    }
}

/// Hash builder for MurmurHash3 32-bit implementation of the 128-bit hashing algorithm.
#[derive(Default)]
pub struct Murmur3Hash128 {}

impl BuildHasher for Murmur3Hash128 {
    type Hasher = Murmur3Hasher128;

    fn build_hasher(&self) -> Self::Hasher {
        Self::Hasher::default()
    }
}

// The code below is adapted from C++ code with the following disclaimer
//-----------------------------------------------------------------------------
// MurmurHash3 was written by Austin Appleby, and is placed in the public
// domain. The author hereby disclaims copyright to this source code.

/// MurmurHash3 32-bit implementation of the 128-bit hashing algorithm.
/// This version allows you to specify a seed.
pub fn hash128_with_seed<T: AsRef<[u8]>>(v: T, seed: u32) -> u128 {
    let data = v.as_ref();
    let n_blocks = data.len() / 16;

    const C1: u32 = 0x239b961b;
    const C2: u32 = 0xab0e9789;
    const C3: u32 = 0x38b34ae5;
    const C4: u32 = 0xa1e38b93;

    const D1: u32 = 0x561ccd1b;
    const D2: u32 = 0x0bcaa747;
    const D3: u32 = 0x96cd1c35;
    const D4: u32 = 0x32ac3b17;

    let mut h1: u32 = seed;
    let mut h2: u32 = seed;
    let mut h3: u32 = seed;
    let mut h4: u32 = seed;

    // body
    for i in 0..n_blocks {
        let mut k1 = get_u32(data, (i + 0) * 4);
        let mut k2 = get_u32(data, (i + 1) * 4);
        let mut k3 = get_u32(data, (i + 2) * 4);
        let mut k4 = get_u32(data, (i + 3) * 4);

        k1 = k1.wrapping_mul(C1);
        k1 = k1.rotate_left(15);
        k1 = k1.wrapping_mul(C2);
        h1 ^= k1;
        h1 = h1.rotate_left(19);
        h1 = h1.wrapping_add(h2);
        h1 = (h1.wrapping_mul(5)).wrapping_add(D1);

        k2 = k2.wrapping_mul(C2);
        k2 = k2.rotate_left(16);
        k2 = k2.wrapping_mul(C3);
        h2 ^= k2;
        h2 = h2.rotate_left(17);
        h2 = h2.wrapping_add(h3);
        h2 = (h2.wrapping_mul(5)).wrapping_add(D2);

        k3 = k3.wrapping_mul(C3);
        k3 = k3.rotate_left(17);
        k3 = k3.wrapping_mul(C4);
        h3 ^= k3;
        h3 = h3.rotate_left(15);
        h3 = h3.wrapping_add(h4);
        h3 = (h3.wrapping_mul(5)).wrapping_add(D3);

        k4 = k4.wrapping_mul(C4);
        k4 = k4.rotate_left(18);
        k4 = k4.wrapping_mul(C1);
        h4 ^= k4;
        h4 = h4.rotate_left(13);
        h4 = h4.wrapping_add(h1);
        h4 = (h4.wrapping_mul(5)).wrapping_add(D4);
    }

    // tail
    //let tail = data.clone();
    let tail_num = n_blocks * 16;
    let mut k1 = 0;
    let mut k2 = 0;
    let mut k3 = 0;
    let mut k4 = 0;
    for i in (1..=data.len() & 15).rev() {
        match i {
            15 => k4 ^= (data[tail_num + 14] as u32) << 16,
            14 => k4 ^= (data[tail_num + 13] as u32) << 8,
            13 => {
                k4 ^= (data[tail_num + 12] as u32) << 0;
                k4 = k4.wrapping_mul(C4);
                k4 = k4.rotate_left(18);
                k4 = k4.wrapping_mul(C1);
                h4 ^= k4;
            }

            12 => k3 ^= (data[tail_num + 11] as u32) << 24,
            11 => k3 ^= (data[tail_num + 10] as u32) << 16,
            10 => k3 ^= (data[tail_num + 9] as u32) << 8,
            9 => {
                k3 ^= (data[tail_num + 8] as u32) << 0;
                k3 = k3.wrapping_mul(C3);
                k3 = k3.rotate_left(17);
                k3 = k3.wrapping_mul(C4);
                h3 ^= k3;
            }

            8 => k2 ^= (data[tail_num + 7] as u32) << 24,
            7 => k2 ^= (data[tail_num + 6] as u32) << 16,
            6 => k2 ^= (data[tail_num + 5] as u32) << 8,
            5 => {
                k2 ^= (data[tail_num + 4] as u32) << 0;
                k2 = k2.wrapping_mul(C2);
                k2 = k2.rotate_left(16);
                k2 = k2.wrapping_mul(C3);
                h2 ^= k2;
            }

            4 => k1 ^= (data[tail_num + 3] as u32) << 24,
            3 => k1 ^= (data[tail_num + 2] as u32) << 16,
            2 => k1 ^= (data[tail_num + 1] as u32) << 8,
            1 => {
                k1 ^= data[tail_num] as u32;
                k1 = k1.wrapping_mul(C1);
                k1 = k1.rotate_left(15);
                k1 = k1.wrapping_mul(C2);
                h1 ^= k1;
            }
            _ => {} // should never occur
        }
    }

    // finalization
    h1 ^= data.len() as u32;
    h2 ^= data.len() as u32;
    h3 ^= data.len() as u32;
    h4 ^= data.len() as u32;

    h1 = h1.wrapping_add(h2).wrapping_add(h3).wrapping_add(h4);
    h2 = h2.wrapping_add(h1);
    h3 = h3.wrapping_add(h1);
    h4 = h4.wrapping_add(h1);

    h1 = fmix32(h1);
    h2 = fmix32(h2);
    h3 = fmix32(h3);
    h4 = fmix32(h4);

    h1 = h1.wrapping_add(h2).wrapping_add(h3).wrapping_add(h4);
    h2 = h2.wrapping_add(h1);
    h3 = h3.wrapping_add(h1);
    h4 = h4.wrapping_add(h1);

    (h1 as u128) << 96 | (h2 as u128) << 64 | (h3 as u128) << 32 | (h4 as u128)
}

/// MurmurHash3 32-bit implementation of the 128-bit hashing algorithm.
/// The seed is always 0 in this version.
pub fn hash128<T: AsRef<[u8]>>(v: T) -> u128 {
    hash128_with_seed(v, 0)
}

#[inline(always)]
fn get_u32(data: &[u8], i: usize) -> u32 {
    let buf = [data[i], data[i + 1], data[i + 2], data[i + 3]];
    u32::from_le_bytes(buf)
}

#[inline(always)]
fn fmix32(h: u32) -> u32 {
    let mut input = h;
    input ^= input >> 16;
    input = input.wrapping_mul(0x85ebca6b);
    input ^= input >> 13;
    input = input.wrapping_mul(0xc2b2ae35);
    input ^= input >> 16;
    input
}

#[cfg(test)]
mod test {
    #[test]
    fn compliance_test() {
        assert_eq!(
            crate::murmur::murmur3_128::hash128_with_seed("StandardBlockEntityDescriptorV4", 4919),
            0x4da5b4125adab9dc7d30c1c10bb975f7
        );
    }
}